This document describes ChemEngine, a program that can extract 3D molecular data from PDF files. ChemEngine uses pattern recognition to identify molecular coordinates in supplementary scientific articles. It generates standard molecular data like bond matrices and atomic coordinates that can then be used for computational analysis. The methodology was demonstrated on three case studies involving different coordinate data formats. ChemEngine accurately extracted coordinates and produced computational results like energies that agreed with original literature values. The tool aims to automate the conversion of molecular data from PDFs into a format suitable for computational workflows.

1. ChemEngine: Harvest Chemical
Data from PDF Files
M. Karthikeyan, Ph.D
CSIR-National Chemical Laboratory
http://moltable.ncl.res.in/
Pune, India

2. ChemEngine

4. ChemEngine
• Digital access to chemical journals resulted in a vast array of
molecular information that is now available in the supplementary
material files in PDF format. However, extracting this molecular
information, generally from a PDF document format is a daunting
task.
• Here we present an approach to harvest 3D molecular data from
the supporting information of scientific research articles that are
normally available from publisher’s resources.
• In order to demonstrate the feasibility of extracting truly
computable molecules from PDF file formats in a fast and efficient
manner, we have developed an application, namely ChemEngine.
• This program recognizes textual patterns from the supplementary
data and generates standard molecular structure data (bond matrix,
atomic coordinates) that can be subjected to a multitude of
computational processes automatically.

5. ChemEngine
• The methodology has been demonstrated via three case studies on
different formats of coordinates data stored in supplementary information
files, wherein ChemEngine selectively harvested the atomic coordinates
and interpreted them as molecules with high accuracy.
• The reusability of extracted molecular coordinate data was demonstrated
by computing Single Point Energies (SPEs) that were in close agreement
with the original computed data provided with the articles.
• It is envisaged that the methodology will enable large scale conversion of
molecular information from supplementary files available in PDF format
into a collection of ready- to- compute molecular data to create an
automated workflow for advanced computational processes.
• Software in the form of jar file is available for downloading at the
Sourceforge site.

6. General Workflow

7. General Workflow Input Data
Format
Pattern
Recognition
Regular
Expression
Generate
Coordinates

8. Pseudo code:
(Co-ordinate Text).matches(“Regular Expression Pattern
with Delimiter Definition”);
For Example: Delimiter: Comma
String_Data.matches("^[A-Za-z0-
9]{1,2},[0]{0,1}[,]{0,1}-{0,1}.{1,2}[0-9]{1,10},-
{0,1}.{1,2}[0-9]{1,10}.{1,}")
Delimiter: Space
String_Data.matches("^[A-Za-z0-
9]{1,2}s+[0]{0,1}[s+]{0,1}-{0,1}.{1,2}[0-9]{0,10}s+-
{0,1}.{1,2}[0-9]{0,10}.{1,}")
Pattern Recognition

9. Harvesting Molecular Data
Bond matrix data
(Computed)
Coordinates Data
(From PDF)
(To be Processed)
Non- Molecular Data
(From PDF)
(To be Excluded)
0:----
Mol_0 1 C1 H2 1.09011 1.200 -0.109889
Mol_0 2 C1 H3 1.08923 1.200 -0.110769
Mol_0 3 C1 H4 1.08923 1.200 -0.110769
Mol_0 4 C1 S5 1.831868 1.55 0.281868
Mol_0 5 S5 H6 1.34383 1.410 -0.066162
C -0.04781100 1.16216400 0.00000000
H -1.09556300 1.46309200 0.00000000
H 0.43082600 1.55738100 0.89506100
H 0.43082600 1.55738100 -0.89506100
S -0.04781100 -0.66970400 0.00000000
H 1.28575000 -0.83557700 0.00000000
S1
SUPPORTING INFORMATION
Thiol-Ene Click Chemistry:
Computational and Kinetic
Analysis of the Influence of Alkene
Functionality.
Brian H. Northrop* and Roderick N.
Coffey
Department of Chemistry
Wesleyan University, Middletown,
Connecticut 06459.

10. AS
C -1.1744 0.5417 -0.9605
S -0.0501 -0.0786 0.1566
C 1.0931 -1.1415 -0.7978
C 1.1877 1.1781 0.7204
H -2.1028 0.8352 -0.4693
H -0.7923 1.2464 -1.7030
6 -1.587291 -1.111365 4.067000
6 -0.971341 0.146218 4.067000
6 0.409526 0.261943 4.067000
7 1.580845 0.365064 4.067000
1 -2.671280 -1.185650 4.067000
1 -1.001643 -2.026790 4.067000
AS
Delimiter
(Space)
AN
Delimiter
(Space)
Delimiter
(Comma)
C,0,0.2760228632,-1.0699067694,1.2573191608
C,0,0.1782250457,1.2145223184,1.2478315394
C,0,0.178225046,1.2145223184,-1.2478315394
C,0,0.2760228634,-1.0699067694,-1.2573191608
C,0,-1.15402211,-0.680015634,-1.5218932206
C,0,-1.2153634407,0.7116626186,-1.5239653892
ChemEngine
Pattern Recognition
C -1.1744 0.5417 -0.9605
S -0.0501 -0.0786 0.1566
C 1.0931 -1.1415 -0.7978
C 1.1877 1.1781 0.7204
H -2.1028 0.8352 -0.4693
H -0.7923 1.2464 -1.7030
C1 S2 1.7019797266712668 1.55 0.1519797266712668
C1 H5 1.0905715244769594 1.2000000000000002 -0.1094284755230408
C1 H6 1.0926613153214495 1.2000000000000002 -0.10733868467855068
X Y ZAS X Y ZAN AS X Y Z
AN X Y Z
Optional
CT
Coordinate matrix
3D MOLECULE

11. Bond Recognition
A1
A2
r1 r2
l1
l1
Bond Recognition
A1 A2
l1
r1’ r2’
l1
A1 A2
r1 r2
d1
0.35 Å
A1 A2
r'1 r‘2
(a)
(b)
(c)

12. Compare the Bond distance
ChemEngine Gaussian 09

13. Computed Conformational Energy
R² = 0.9979
-1200
-1000
-800
-600
-400
-200
0
-1200 -1000 -800 -600 -400 -200 0
Energy(RHF)
Energy (CBS-QB3)

14. Case Studies
Entry Case Study N= molecules Regular Expression pattern Format
&
Delimit
er
1 Epoxide formation
from sulfur ylides and
aldehydes
29 ^[A-Za-z0-9]{1,2}s+-
{0,1}.{1,2}[0-9]{1,8}s+-
{0,1}.{1,2}[0-9]{1,8}.{1,}
PDF
Space
2 Thiol ene click chemistry 115 ^[A-Za-z0-9]{1,2}s+-
{0,1}.{1,2}[0-9]{1,8}s+-
{0,1}.{1,2}[0-9]{1,8}.{1,}
Text
Space
3 Design of
tetra(arenediyl)bis(allyl)
derivatives for cope
rearrangement
transition states
55 ^[A-Za-z0-
9]{1,2},[0]{0,1}[,]{0,1}-
{0,1}.{1,2}[0-9]{1,10},-
{0,1}.{1,2}[0-9]{1,10}.{1,}
PDF
Comma

15. ChemEngine

16. Harvesting Molecular Data

17. Thank You